SCIEN Talk

Conventional cameras record all light falling onto their sensor regardless of its source or its 3D path to the camera. In this talk I will present a emerging family of coded-exposure video cameras that can be programmed to record just a fraction of the light coming from an artificial source---be it a common street lamp or a programmable projector---based on the light path's geometry or timing. Live video from these cameras offers a very unconventional view of our everyday world in which refraction and scattering can be notice with the naked eye can become apparent, and the flicker of electric lights can be turned into a powerful cue for analyzing the electrical grid from room to city.

I will discuss the unique optical properties and power efficiency of these "transport aware cameras" through three case studies: the ACam for analyzing the electrical grid, EpiScan3D for robust progress toward designing a computational CMOS sensor for coded two-bucket imaging---a novel capability that promises much more flexible and powerful transport-aware cameras compared to existing off-the-shelf solutions.

Human binocular vision and acuity, and the accompanying 3D retinal processing of the human eye and brain are specifically designed to promote situational awareness and understanding in the natural 3D world. The ability to resolve depth within a scene whether natural or artificial improves our spatial understanding of the scene and as a result reduces the cognitive load accompanying the analysis and collaboration on complex tasks.

A light-field display projects 3D imagery that is visible to the unaided eye (without glasses or head tracking) and allows for perspective correct visualization within the display's projection volume. Binocular disparity, occlusion, specular highlights and gradient shading, and other expected depth cues are correct from the viewer's perspective as in the natural real-world light-field.

Light-field displays are no longer a science fiction concept and a few companies are producing impressive light-field display prototypes. This presentation will review:
· The application agnostic light-field display architecture being developed at FoVI3D.
· General light-field display properties and characteristics such as field of view, directional resolution, and their effect on the 3D aerial image.
· The computation challenge for generating high-fidelity light-fields.
· A display agnostic ecosystem.

Demo after the talk: The FoVI3D Light-field Display Developer Kit (LfD DK2) is a prototype, wide field-of-view, full parallax, monochrome light-field display capable of projecting ~100,000,000 million unique rays to fill a 9cm x 9cm x 9cm projection volume. The particulars of the light-field compute, photonics subsystem and hogel optics will be discussed during the presentation.

New methods for capturing highly detailed models of moving real world scenes with cameras, i.e., models of detailed deforming geometry, appearance or even material properties, become more and more important in many application areas. They are needed in visual content creation, for instance in visual effects, where they are needed to build highly realistic models of virtual human actors. Further on, efficient, reliable and highly accurate dynamic scene reconstruction is nowadays an important prerequisite for many other application domains, such as: human-computer and human-robot interaction, autonomous robotics and autonomous driving, virtual and augmented reality, 3D and free-viewpoint TV, immersive telepresence, and even video editing.

The development of dynamic scene reconstruction methods has been a long standing challenge in computer graphics and computer vision. Recently, the field has seen important progress. New methods were developed that capture - without markers or scene instrumentation - rather detailed models of individual moving humans or general deforming surfaces from video recordings, and capture even simple models of appearance and lighting. However, despite this recent progress, the field is still at an early stage, and current technology is still starkly constrained in many ways. Many of today's state-of-the-art methods are still niche solutions that are designed to work under very constrained conditions, for instance: only in controlled studios, with many cameras, for very specific object types, for very simple types of motion and deformation, or at processing speeds far from real-time.

In this talk, I will present some of our recent works on detailed marker-less dynamic scene reconstruction and performance capture in which we advanced the state of the art in several ways. For instance, I will briefly show new methods for marker-less capture of the full body (like our VNECT approach) and hands that work in more general environments, and even in real-time and with one camera. I will then show some of our work on high-quality face performance capture and face reenactment. Here, I will also illustrate the benefits of both model-based and learning-based approaches and show how different ways to join the forces of the two open up new possibilities. Live demos included!

Cyber-physical/human systems (CPS/CHS) are set to play an increasingly visible role in our lives, advancing research and technology across diverse disciplines. I am exploring novel synergies between three emerging CPS/CHS technologies of prospectively broad societal impact, virtual/augmented reality (VR/AR), the Internet of Things (IoT), and autonomous micro-aerial robots (UAVs). My long-term research objective is UAV-IoT-deployed ubiquitous VR/AR immersive communication that can enable virtual human teleportation to any corner of the world. Thereby, we can achieve a broad range of technological and societal advances that will enhance energy conservation, quality of life, and the global economy.
I am investigating fundamental problems at the intersection of signal acquisition and representation, communications and networking, (embedded) sensors and systems, and rigorous machine learning for stochastic control that arise in this context. I envision a future where UAV-IoT-deployed immersive communication systems will break existing barriers in remote sensing, monitoring, localization and mapping, navigation, and scene understanding. The presentation will outline some of my present and envisioned investigations. Interdisciplinary applications will be highlighted.

'Image volumes' refer to realizations of images in other dimensions such as time, spectrum, and focus. Recent advances in scientific, medical, and consumer applications demand improvements in image volume capture. Though image volume acquisition continues to advance, it maintains the same sampling mechanisms that have been used for decades; every voxel must be scanned or captured in parallel and is presumed independent of its neighbors. Under these conditions, improving performance comes at the cost of increased system complexity, data rates, and power consumption.

This talk describes systems and methods with which to efficiently detect and visualize image volumes by temporally encoding the extra dimensions' information into 2D measurements or displays. Some highlights of my research include video and 3D recovery from photographs, and true-3D augmented reality image display by time multiplexing. In the talk, I show how temporal optical coding can improve system performance, battery life, and hardware simplicity for a variety of platforms and applications.

Computer-graphics engineers and vision scientists want to generate images that reproduce realistic depth-dependent blur. Current rendering algorithms take into account scene geometry, aperture size, and focal distance, and they produce photorealistic imagery as with a high-quality camera. But to create immersive experiences, rendering algorithms should aim instead for perceptual realism. In so doing, they should take into account the significant optical aberrations of the human eye. We developed a method that, by incorporating some of those aberrations, yields displayed images that produce retinal images much closer to the ones that occur in natural viewing. In particular, we create displayed images taking the eye's chromatic aberration into account. This produces different chromatic effects in the retinal image for objects farther or nearer than current focus. We call the method ChromaBlur. We conducted two experiments that illustrate the benefits of ChromaBlur. One showed that accommodation (eye focusing) is driven quite effectively when ChromaBlur is used and that accommodation is not driven at all when conventional methods are used. The second showed that perceived depth and realism are greater with imagery created by ChromaBlur than in imagery created conventionally. ChromaBlur can be coupled with focus-adjustable lenses and gaze tracking to reproduce the natural relationship between accommodation and blur in HMDs and other immersive devices. It can thereby minimize the adverse effects of vergence-accommodation conflicts.

Between ever increasing pixel counts, ever cheaper sensors, and the ever expanding world-wide-web, natural image data has become plentiful. These vast quantities of data, be they high frame rate videos or huge curated datasets like Imagenet, stand to substantially improve the performance and capabilities of computational imaging systems. However, using this data efficiently presents its own unique set of challenges. In this talk I will use data to develop better priors, improve reconstructions, and enable new capabilities for computational imaging systems.

In this talk, I will present two recently published papers at the annual SIGGRAPH ASIA 2017. For the first paper, we present a 4D light field sampling and rendering system for light field displays that can support both foveation and accommodation to reduce rendering cost while maintaining perceptual quality and comfort. For the second paper, we present a light field based Computer Generated Holography (CGH) rendering pipeline allowing for reproduction of high-definition 3D scenes with continuous depth and support of intra-pupil view dependent occlusion using computer generated hologram. Our rendering and Fresnel integral accurately accounts for diffraction and supports various types of reference illumination for holograms.

A brief overview of Street View from it's inception 10 years ago until now will be presented. Street level Imagery has been the prime objective for Google's Street View in the past, and has now migrated into a state-of-the-art mapping platform. Challenges and solutions to the design and fabrication of the imaging system and optimization of hardware to align with specific software post processing will be discussed. Real world challenges of fielding hardware in 80+ countries will also be addressed.

Many vision tasks require not just categorizing a well-composed human-taken photo, but also intelligently deciding "where to look" in order to get a meaningful observation in the first place. We explore how an agent can anticipate the visual effects of its actions, and develop policies for learning to look around actively---both for the sake of a specific recognition task as well as for generic exploratory behavior. In addition, we examine how a system can learn from unlabeled video to mimic human videographer tendencies, automatically deciding where to look in unedited 360 degree panoramas. Finally, to facilitate 360 video processing, we introduce spherical convolution, which allows application of off-the-shelf deep networks and object detectors to 360 imagery.